4f96c63 emit_x64_vector_floating_point: Simplify FPVector{Min,Max}
e15fdfe emit_x64_vector_floating_point: Simplify Get*Vector functions
734a00b emit_x64_floating_point: Remove EmitProcessNaNs
fd45191 devirtualize: Replace DEVIRT macro with function template
67ba5d0 fuzz_with_unicorn: Remove FCVT_float from ignore list
66e6dd1 a32_emit_x64: std::move A32::UserConfig in the constructor
b4890b6 emit_x64_floating_point: Use EmitPostProcessNaNs in EmitFPMulX
18b2943 emit_x64_floating_point: Remove unnecessary DenormalsAreZero from EmitFPSingleToDouble and EmitFPDoubleToSingle
df1f81f emit_x64_floating_point: Simplify EmitFP{Min,Max}{,Numeric}{32,64}
21fb1c3 emit_x64_floating_point: Reduce NaN processing overhead
f5c9f0f A64: Implement FMULX, scalar single/double variant
8f47773 IR: Implement FPMulX IR instruction
79e6440 fuzz_with_unicorn: Randomize SP
33c80e3 fuzz_with_unicorn: Randomize PC
8d41024 testenv: Make code_mem mobile
a9fae0e emit_x64_vector: Vectorize 32-bit variants of paired min/max
8926a92 emit_x64_vector: Improve code emission of VectorGetElement* for index == 0
e20bd38 reg_alloc: Do a UseScratch if a Use destination is too small
a19fa0e fuzz_with_unicorn: Randomize FPCR.AHP and FPCR.FZ16
775f368 emit_x64_floating_point: AVX implementation of ForceToDefaultNaN
71018a1 emit_x64_vector_floating_point: Prefer blendvp{s,d} to vblendvp{s,d} where possible
137f4b3 backend_x64: Remove all use of xmm0
e73d67a emit_x64_vector_floating_point: AVX implementation of ForceToDefaultNaN
43cca54 emit_x64_vector_floating_point: Reduce codesize of ForceToDefaultNaN
5dc40f4 emit_x64_vector_floating_point: Reduce codesize of EmitTwoOpVectorOperation
07622ee emit_x64_vector_floating_point: Correct FMA in FTZ mode
621c85b emit_x64_floating_point: DenormalsAreZero is redundant as hardware already does DAZ
3d0ebaa emit_x64_floating_point: FlushToZero is redundant as hardware already does FTZ
f626ff8 backend_x64: Fix FPVectorMulAdd and FPMulAdd NaN handling with denormals
adeb9d9 a32/fuzz_arm: Disable vfp tests
19ea70d fuzz_with_unicorn: Randomize FPCR.FZ
895db36 backend_x64: Fix bugs when FPCR.FZ=1
d7e2de2 fuzz_with_unicorn: Extract RandomFpcr function
c858d6c fp/info: Deduplicate functions
5b88ec2 emit_x64_floating_point: Deduplicate EmitFPMulAdd implementation
The benefit of std::string_view comes from the idea of avoiding copies
(essentially acting as a non-owning view), however if we're just going
to copy into a local variable immediately, there's not much benefit
gained here.
Ensures both operands have the same sign in the comparison.
While we're at it, we can get rid of the redundant casting of ub_size to
an int. This type will always be trivial and alias a built-in type (not
doing so would break backwards compatibility at a standard level).
All calling code assumes that the rasterizer will be in a valid state,
which is a totally fine assumption. The only way the rasterizer wouldn't
be is if initialization is done incorrectly or fails, which is checked
against in System::Init().
We move the initialization of the renderer to the core class, while
keeping the creation of it and any other specifics in video_core. This
way we can ensure that the renderer is initialized and doesn't give
unfettered access to the renderer. This also makes dependencies on types
more explicit.
For example, the GPU class doesn't need to depend on the
existence of a renderer, it only needs to care about whether or not it
has a rasterizer, but since it was accessing the global variable, it was
also making the renderer a part of its dependency chain. By adjusting
the interface, we can get rid of this dependency.
This amends cases where crashes can occur that were missed due to the
odd way the previous code was set up (using 3DS memory regions that
don't exist).
Using member variables for referencing the segments array increases the
size of the class in memory for little benefit. The same behavior can be
achieved through the use of accessors that just return the relevant
segment.
